Lyft is teaming up with AWS and Anthropic to launch a new agentic AI system designed to make rider and driver support feel immediate and intuitive.
The upgrade replaces rigid menus with AI that can read context, understand issues, and move quickly to resolve them.
In Lyft’s announcement, the company said the Claude-powered system running on Amazon Bedrock is already reshaping customer support for its rider and driver community, with more AI-driven features on the way.
Support that reads the room — and your recent rides
Lyft’s new intent agent jumps in as soon as a message comes through, pulling up recent trips, payment details, or past issues to zero in on what needs fixing.
Ask why your earnings aren’t showing up, and the system already knows you wrapped three rides minutes ago and can take the next step to resolve it. If a rider mentions a missing wallet, the AI connects the dots to the last completed trip and initiates the recovery flow without sending them through a category maze.
This is what makes the system agentic: instead of responding with static answers, it uses Claude to understand context, surface relevant backend data, and carry out tasks that once required multiple screens of menus and retries. It’s a shift from tapping through hundreds of options to having support that feels like it’s already in the passenger seat.
Proof that riders and drivers are embracing the new system
Lyft says the agentic AI is already handling thousands of support requests each day, with more than half resolved in under three minutes. Average resolution times have dropped by a whopping 87%, which the company credits to the AI tool that can interpret intent and act without sending users through repeat prompts.
Adoption is climbing as well. Driver usage of the AI surged by 70% in 2025, showing how quickly the tool is becoming part of their daily workflow. The assistant is available around the clock in English and Spanish, providing Lyft with a support layer that can keep up with the volume of its 40 million riders and one million drivers without slowing down.
Ameena Gill, who oversees safety and customer care at Lyft, said the launch reflects how “the future of customer service has fundamentally shifted through AI.”
How months of groundwork set up this moment
Long before the new system reached riders and drivers, Lyft laid out a plan with Anthropic to co-develop AI products, gain early access to next-generation model capabilities, and train its engineering teams to build with them. The goal was to create tools that could understand nuance, respond with a human tone, and operate safely at scale, a foundation that would later support more advanced features across the platform.
The agentic support system is the first major result of that groundwork. Lyft noted months of joint development and early model testing with Anthropic gave its teams the runway to build an assistant that can interpret intent and take meaningful action.
The company said it will share its early learnings at AWS re:Invent this week as it continues building out its customer-first AI approach.
Inside Amazon, a growing group of employees is publicly challenging the company’s fast-moving AI strategy.